parallelism of datum E wrt to datum B

[[K...]]

I measured flatness of datum B as 0.024mm, datum E 0.103mm and analyzed highest and lowest points of datum E which are 21.922 and 22.046.

When I callout parallelism of datum E wrt to datum B, I got 0.175mm. I don't understand why I am getting that?

I really appreciate if any one resolve this issues.

[[Da...]]

While flatness can have an effect on parallelism you could have 2 perfectly flat planes that are not parallel. Parallelism shows a taper or angle or slope of 1 plane in relation to the other plane. You can check it manually by setting the part on a flat surface with datum B contacting the surface and use a height gage and indicator to sweep the E surface. This will show you where it is out or if it may have moved on the CMM when it was being checked.

[[Da...]]

Also, the flatness of the datum plane B is not part of the result, since depending on your settings (ISO 5459) it will either be a tangent plane or a Gaussian plane.

Well, in a way, when the datum plane is calculated as a tangent plane, the flatness does matter, of course, but only since it changes the orientation of the datum plane, the flatness itself will not be added to the parallelism. That’s why you should always use filter settings and outlier elimination when using the ISO 5459 setting on datum planes.

[[An...]]

___

[[Da...]]

It’s always the question: min or max to what? I don’t know where you pulled the max and min point of E from, but they usually will be the min/max points of the feature itself (so in the direction of E). The flatness is always calculated to the minimum zone element and therefore always smaller than the the difference of the max to the min point. Both values are not feasible to compare them to the parallelism value, because the direction in which they were calculated is wrong, and parallelism is all about the direction or orientation of something.

In the parallelism evaluation the difference from the high point and the low point of E will be calculated, but in the direction of the datum B, and that direction will change whether you use a Gaussian B or an outer tangential element (properly filtered of course), as it should be here.

In ASME and in ISO, there will also be different methods to establish the outer tangential element. In ISO it‘s basically a tangential plane parallel to the Minimum Zone Element, in ASME there are different methods to create such a plane, the preferred is an L2 plane (Filtered Gaussian tangential).

[[K...]]

Thanks for the input. just for info. I contracted the plain using 30 hits with LSQ evaluation method ( and also tried outer Tangential Element ) .

Each point is equally distributed at 12 degree angles.

Daniel: Can I use a filter and outliers option when constructing the plain using 30 hits?

Thank you

[[Da...]]

Please sign in to view this quote.

Filters won’t work with too few points and I think Calypso won’t eliminate outliers if it would mean to delete more than 10 % of the total points, which would be only three when taking 30 hits.

[[Er...]]

With so few points try L1 Evaluation. It is not so sensitive to outliers.

When taking points for -B-, try to make them as close to the edge as possible to make the calculated plane larger. Maybe this will help.

Maybe try this? See if the numbers agree now?
Cartesian Distance from plane to point. Set evaluation for plane to OTE.

The parallelism is calculated from the calculated centerline of the -E- plane. So points at imperfections in the surface of -E- would make parallelism larger than the variation.

An old school granite check was to put a piece of flat ground steel on top and measure to eliminate the surface variation. Not ISO standard, but ballpark.